designing new trike lights

[charlie_r]

Thanks to this highly informative forum, I now know my own lack of knowledge in lighting design. For this, I do thank all of the more knowledgeable folks here, especially Scheinwerfermann.

What I am looking to do is build a better lighting system for my work trike. What I have now I know to be highly inadequate, and even dangerous to myself and others, due fully to waaaaaay too much glare into approaching auto drivers and pedestrians faces.

Bearing in mind that I don't use the headlights on a regular basis, I still would like something a bit more safe for use in dark conditions.

Here is a pic of the current light bar:

(clickable)

All units except high beam are 1W gull wing dies on series wired triple aluminum PCB, high beam are 3W.

Low beams are driven at 150mA, high beam at 600mA. Turn signals are amber, at 350mA. Power comes from a 12V AGM battery.

The triple optics are supposedly 15*, but I have my doubts. A lot of spread in all directions, far greater than I would expect from these. As mounted, there is no way to adjust the aim.

As I do have a limited budget, low cost is a concern.

I don't particularly like the "bike mounted flashlight" look, and don't have the budget for commercially made bike lights.

What I am looking to do is separate the low, high, and turn signals so each type can be aimed more appropriately for street use, without blinding other road users.

I know nothing about using aspherical optics, should that be needed for this, but I am willing to learn.

This is what the lights are to be mounted on:



Suggestions anyone?

Charlie

 

[Steve K]

do you have specific design requirements for each of the functions (low beam, high beam, and turn signal)? Something along the line of beam width in the horizontal and vertical direction, lumens of output, input power and voltage, cost, etc?

To get a fairly tight headlight beam, I'm pleased with the Ledil Rocket Smooth Spot....
http://www.digikey.com/product-detail/en/CA10324_ROCKET-SS/711-1053-ND/1971711

 

[charlie_r]

Steve,

Beam width, horizontal -- 10-12 ft wide, enough to cover 1 lane of traffic with some curb spill. Vertical, low beam -- 40-50 ft in front of the trike. I don't know really how far I would need there, as my speed is typically under 20 mph. High beam, probably not really needed as all of my current and rare night trips are in town with fair to poor street lights.

Turn signals will need to be done similar to the rear lights, for greater side visibility.











Voltage and current will be provided by a 12V 10Ah AGM battery. Budget constraints do not currently allow for even a decent set of NiMH cells, let alone what I really want, LiFePo.

Burn time -- I'm aiming for 3-4 hours. I expect to have to go to PWM to realize this.

Lumens output -- I'm not sure just how high I need to go to attain the goal of "see and be seen". Surely not "surface of the sun" bright, but high enough to see, and get other road users attention without blinding them as happens now.

My main concern with needing to make this change is as the lights are now, they are dangerous to oncoming traffic. I don't want my lights to be the reason someone gets seriously injured or worse because oncoming drivers couldn't see the pedestrian or bike in their lane. My ignorance of good lighting principle shouldn't be the cause of other's injuries.

The lens you linked to. Can it be used with the round button gull wing LEDs? Or would I need to go with XP-G or similar? I ask, because I have more than enough of the older 1W & 3W LEDs on hand to complete the project.

Thank you!

Charlie

 

[Steve K]

Steve,

Beam width, horizontal -- 10-12 ft wide, enough to cover 1 lane of traffic with some curb spill. Vertical, low beam -- 40-50 ft in front of the trike. I don't know really how far I would need there, as my speed is typically under 20 mph. High beam, probably not really needed as all of my current and rare night trips are in town with fair to poor street lights.

well, I guess I was trying to guide you towards calculating what sort of beam angle you might want. Admittedly, I haven't done it that way myself.. I seem to have made my best guess about beam width, and then adjusted the selection. The Ledil optic that I linked to has a 8 degree "full width half max" beam angle. This means that it is 4 degrees from the center to the points where the brightness has dropped to half of the max value.

My main concern with needing to make this change is as the lights are now, they are dangerous to oncoming traffic. I don't want my lights to be the reason someone gets seriously injured or worse because oncoming drivers couldn't see the pedestrian or bike in their lane. My ignorance of good lighting principle shouldn't be the cause of other's injuries.

The lens you linked to. Can it be used with the round button gull wing LEDs? Or would I need to go with XP-G or similar? I ask, because I have more than enough of the older 1W & 3W LEDs on hand to complete the project.

The linked site says it is intended for Cree XR-C and XR-E leds. I use them with XR-E's. If you have leds that you want to use. you'll want to check with the various optics manufacturers to see what optics are compatible and what the beam angles are. If you find something that looks good, I'd just buy one or two and see if the beam angle is what you are looking for. With a tighter beam angle, you'll be able to use less power (and leds) and produce the same illumination of the road surface.

 

[Esko]

My main concern with needing to make this change is as the lights are now, they are dangerous to oncoming traffic. I don't want my lights to be the reason someone gets seriously injured or worse because oncoming drivers couldn't see the pedestrian or bike in their lane. My ignorance of good lighting principle shouldn't be the cause of other's injuries.

In other words, you want a cut-off beam. There is at least one easy way to do it. You can use reflectors of your choice, shave the upper parts away and replace the shaven parts with a straight mirror surface. This way.

 

[charlie_r]

Esko, Thanks for that!

I hadn't realized that was an option. Is it possible to polish the cut surface to the point where it will again reflect without having to resort to some sort of mirror material? Yes, I've been reading about the TIR phenomenon (Total Internal Reflection). I had expected that any optics would need mirroring on the outside to be used for lighting. "Google is my friend".

I will begin experimenting with this idea, while I search for eliptical optics that will work with Luxeon III/SSC P4/Edison Edixeon emitters. Initial searching yesterday showed that most suppliers are no longer carrying for these LEDs. I may have to go ahead and spend $$ on XP-G's or similar, just to be able to buy optics I want.

Charlie

 

[Esko]

Esko, Thanks for that!

I hadn't realized that was an option. Is it possible to polish the cut surface to the point where it will again reflect without having to resort to some sort of mirror material? Yes, I've been reading about the TIR phenomenon (Total Internal Reflection). I had expected that any optics would need mirroring on the outside to be used for lighting. "Google is my friend".

I will begin experimenting with this idea, while I search for eliptical optics that will work with Luxeon III/SSC P4/Edison Edixeon emitters. Initial searching yesterday showed that most suppliers are no longer carrying for these LEDs. I may have to go ahead and spend $$ on XP-G's or similar, just to be able to buy optics I want.

Charlie

I have never seen anything like that done with a TIR lens, but why not. At least I am sure it could be polished well enough. However, I suggest you to find out more about how total internal reflection works. It all depends on the angle that the light hits the surface. If it comes pretty straight, it will go through. If it comes in small angle, it is reflected. And if it is something in between, part of the light goes through and the rest is reflected. If you make a straight cut like you would do in reflector, the angle changes a lot and most of the light will probably go through the cut surface. A mirror wouldn't necessarily help either, because it reflects the light back inside the lens where it would be reflected multiple times. In worst case scenario it would reflect big part of the light back to the led, making it run hotter.

Taking a small slice away might work (not a big straight horisontal cut). The bad news is that you need to polish the cut surface every time, otherwise you don't know how the effect turned out. You also need to be very careful with your TIR lenses. Grease (from fingers) and tiny little scratches will easily ruin the TIR effect and make the lenses perform worse. Anyway, if you start to experiment with this, be sure to tell us about your findings! :thumbsup:

 

[charlie_r]

Thanks for the encouragement!

From what little I've learned about TIR, I figured that an angle cut starting right at the edge of the LED recess and traversing up at an angle to the front edge would give the best effect. But what angle? That will have to be determined by experimentation, as you suggest.

Since I am going into somewhat new territory (linked thread included - at least what I've read of it so far) I am planning on doing many beam shots with what passes for a camera (point and shoot panasonic Lumix LZ7) and posting the results with rough measurements at each stage.

Currently on hand I have about a dozen 15* lenses that I can experiment with.

I'll probably need to find a local source of jeweler's rouge for polishing the cuts. A thought here: If I leave a polished fresnel effect on the cut, maybe that will widen the beam in the horizontal direction? Watch this space for the answer! :lolsign:

:thanks:

Charlie